Inhalation Exposure and Intake Dose Model
Improvements

Stephen Graham1, Thomas McCurdy1, Janet Burke1, John Langstaff2, Harvey Richmond2,

1ORD/NERL/Human Exposure and Atmospheric Sciences (HEASD)/Exposure Modeling Research Branch (EMRB)

2Office of Air Quality Planning and Standards (OAGPS)/ Air Quality Strategies and Standards Division (AQSSD)/Health and Ecosystems Effects Group (HEEG)

Abstract

EMRB scientists have been improving its human exposure/intake dose model (i.e., Stochastic Human Exposure and Dose Simulation [SHEDS] model) by performing high-priority human exposure assessment research.
One such area of research is in enhancing^ojDuiation-based inhalation exposure and dose modeling, ana the EMRB staff have developed ^ coordinated modeling program to improve methodologies and algorithms utilized

within various exposure models in tl

ngprogr

^ program offices. For example, the OAQPS uses a set of regulatory models, including the Air Pollutants Exposure (APlzX) model (which is the current inhalation model for the
Total Risk Integrated Methodology for Exposure (TRIM.Expo)), and the Hazardous Air Pollutants Exposure Model (HAPEM) to evaluate alternative national ambient air quality standards (NAAQS) and emission standards for
toxic, hazardous air pollutants (HAPS). While the SHEDS model is most similar in structure and function to the APEX model, all three models (as well as others) use the EMRB's Consolidated Human Activity Database
(CHAD) as their source of human activity data. Since CHAD data are fundamental to both the NERL's and OAQPS's exposure models, the EMRB staff have been evaluating the validity and effectiveness of the CHAD to
address significant research questions posed by external scientific review groups. Results of the first three evaluation efforts have been published in peer-reviewed journals; subsequently, the OAQPS is modifying its
_. 				x	x	*=._ .... same jS true for the new 2000 US Census commuting data that the EMR" '	-	-	-

exposure modeling approach to account for the EMRB research findings.'

same is true for the new 2000 US Census commuting data that the EMRB has obtained and modified; it currently is being input into the OAQPS

		------- "	- en consumption relationships that are an integral part of the intake dose-estimating algorithms used in all

the EMRB in coordination with scientists in the OAQPS and provides insight into how these products are

models to replace the 1990 version. In addition, EMRB scientists have developed new ventilation (breathingVto-oxygen consumption relationships that are an integral part of the intake dose-estimating algorithms used in all

of the models. This presentation highlights recent human exposure model improvements and products developed by the EN,nn	"—	»- ~ — -¦ :j- - =--*-•¦ - - -»—«

used by the OAQPS in its regulatory exposure models.

Although this work was reviewed by EPA and approved for publication, it may not necessarily reflect official Agency policy.

¦ Uptake dose is determined by where
people are and what they are doing.

Research Drivers

¦ Exposure/dose metrics should be
consistent with health effect of concern.

M 30
3=25
| 20

Time-Integrated

Inhalation rates are estimated
considering physiological relevance.

BMR * METS
CHAD

RQ/P aC02 "

U

(EE V02)

¦U



Research Project 1:

Commuter Database Development

Houston Example

ISSUE

People are not stationary over
time. Commuting and other
geographic relocations are
important factors in better
estimating inhalation
exposure. APEX and SHEDS
models use 1990 commuting
data and 2000 Census tracts.

SOLUTION
Two databases consisting of
the number of people living in
one tract and commuting to
other tracts were designed to
replace the outdated APEX
and SHEDS-AirToxics
databases for those
commuting <120 km.

FUTURE RESEARCH
A third database was
developed for creation of new
com m uti n g d atab ases that
contain additional attributes
such as:

1.	Separation of individuals
working at home from
those commuting within
their home tract.

2.	Ability to identify home tracts related to a specific work tract.

3.	Inclusion of individuals commuting over 120km, those
working outside the US, and those who work at unknown
locations.

Tract-To-Tract Rows Less Than 120km

Sjti ftmte a l frKuei Wtak

Research Project 2:

CHAD Data Evaluation

ISSUE

CHAD (McCurdy et al., 2000) is a valuable tool for use in human
exposure modeling; however, due to the cross sectional data
comprising CHAD, an in-depth evaluation of why people spend
time in certain microenvironments was necessary to develop
strategies for longitudinal time-location-activity diary structure.

SOLUTION

1.	Develop typology of
explanatory metrics of
human activity for a
specific cohort (McCurdy
and Graham, 2003).

2.	Test typology and
determine significant
factors affecting time spent
in microenvironments for
all individuals (Graham and
McCurdy, 2004)

• Age, gender, PAI, day-
type, temperature
within seasons

3.	Estimate number of diaries
needed to represent inter-
and intra-individual
variability (Xue et al., 2004)

FUTURE RESEARCH	speTtoi

Evaluate exercise levels for children in CHAD to determine if the
types and amount contained in it conform to exercise surveys
recently undertaken by other organizations. If the results are
positive, we will gain increased confidence in usefulness of CHAD
regarding this important attribute.

d

Literature Referenced

Graham SE and McCurdy T. 2004. Developing meaningful cohorts for human exposure models. J Expos Anal Environ Epidemiol. 14: 23-43.
Graham SE and McCurdy T. 2005. Revised Ventilation Rate (VE) equations for use in inhalation-oriented exposure models. US EPA, Washington
DC, EPA/600/X-05/008.

McCurdy T, Glen G, Smith L, and Lakkadi Y. 2000. The National Exposure Research Laboratory's Consolidated Human Activity Database. J

Expos Anal Environ Epidemiol. 10: 566-578.

McCurdy T and Graham S. 2003. Using human activity data in exposure models: analysis of discriminating factors. J Expos Anal Environ
Epidemiol. 13:294-317.

McCurdy T and Graham S. 2004. Analyses to understand relationships among physiological parameters in children and adolescents aged 6-16.

US EPA, Washington, DC, EPA/600/X-04/092.

McCurdy T and Xue J. 2005. Meta-analysis of physical activity index data for US children and adolescents. J Child Health (accepted).

Xue J, McCurdy T, Spengler J and Ozkaynak H. 2004. Understanding variability in time spent in selected locations for 7-12-year old children. J
Expos Anal Environ Epidemiol. 14: 222-233.

Research Project 3:
Ventilation Algorithm

ISSUE

OAQPS requested that EMRB
review the literature on
estimating alveolar ventilation
(VA) since a previous review of
the algorithms used in
pNEM/CO indicated that a
constant in the equation
possibly varied non-l in early
with exercise rate. As an
outgrowth of this work EMRB
decided to first investigate a VE
algorithm for use in both the
APEX and SHEDS inhalation
modules.

SOLUTION

Age (A) and gender (G) were used as independent variables
along with body mass normalized (BM) V02 in a multiple linear
regression (Graham and McCurdy, 2005).

Ln(VE/BM), = b0+(b/Lne.'OI®Ml))+(b!*Ln(A,))+(b3 . G|)+ ^ ^

Parameter estimates (b(), coefficient standard errors (se), and residual
distributions standard deviation estimates (e,) assuming above equation.

Age
(n)

b0
(se)

bi
(se)

b2

(se)

b3

(se)

eb

ew

Rz

<20
(1085)

4.43
(0.06)

1.09
(0.01)

-0.28
(0.01)

0.05
(0.004)

0.10

0.11

0.92

20-<34
(3646)

3.57
(0.08)

1.17
(0.01)

0.11
(0.02)

0.04
(0.003)

0.12

0.13

0.89

34-<61
(1083)

3.19
(0.13)

1.12
(0.01)

0.18
(0.03)

0.04
(0.01)

0.13

0.12

0.89

61+

(457)

2.45
(0.36)

1.04
(0.02)

0.27
(0.08)

-0.03
(0.01)

0.11

0.07

0.89

All coefficient

statistically significan

at p< 0.01











FUTURE RESEARCH

The goal of the research, recognizing that air pollutants will vary in
absorption location depending on the substances' physical and
chemical characteristics, is to have a unified approach for the
various ventilation metrics with V02 at the core of the algorithm. A
method for estimating VA to remain consistent with the VE
estimation is under investigation. Previously, the pathway from
V02 to VA was considered as a linear proportionality (i.e., 19.63)
and also estimated independently from VE. Research indicates
the approximation is reasonable for low to moderate exercise
levels, but there is variability in VA at all exercise levels that are
not accounted for by the point estimate used to modify V02.

-------